Comorbidities Drive Deaths in Low-Risk Pancreatic Cysts

In patients with low-risk mucinous pancreatic cysts, overall mortality was associated with comorbidity burdens quantified with baseline age-adjusted Charlson Comorbidity Index (age-CCI) scores, which may therefore be incorporated into individualized patient surveillance strategies, researchers said.

The analysis, published online in Clinical Gastroenterology and Hepatology, looked at the comorbidity burden and mortality outcomes in a surveillance population with low-risk pancreatic cysts. James J. Farrell, MD, MBChB, of Yale School of Medicine in New Haven, Connecticut, and colleagues found baseline age-CCI had good predictive capacity for 4-year extra-pancreatic mortality, with an area under the curve of 0.87 (P<.00001)

The researchers stratified 440 patients into low- and high-comorbidity groups with an age-CCI cutoff of 4. Not surprisingly, significantly higher overall mortality was observed in the high versus the low CCI group.

Based on regression analysis, the following comorbidities elevated odds ratios for extra-pancreatic mortality (P<0.0001):

• Liver disease by 2.3-8.1
• Solid malignancies by 4.9
• Hematologic malignancies, by 4.7
• Metastatic malignancies by 21.6

“Our study supports the individualized approach and continual assessment of surveillance appropriateness among pancreatic cyst population,” Farrell and colleagues wrote. “Low-risk cyst patients with competing mortality risks will benefit from a shared decision of surveillance cessation/modification.”

According to recent guidelines, cross-sectional and/or endoscopic ultrasound imaging can detect characteristics associated with malignant progression, and this in turn helps determine the appropriateness of surgical resection for patients at high risk of an advanced pathology or nonsurgical surveillance for those at lower risk.

In this setting, CCI has been proposed as a tool to determine comorbidity burden and guide management for patients with mucinous pancreatic cysts (intrapapillary mucinous and mucinous cystic neoplasms), but has not been well studied in lesions without worrisome features and high-risk stigmata.

Study details

The study placed 502 individuals undergoing imaging at a single center in 2016 under prospective surveillance, including 440 with low-risk suspected or presumed mucinous cysts. It excluded 50 and 12 individuals with worrisome features or high-risk stigmata, respectively.

The median overall age in the cohort was 68, compared with 73 in the high-CCI group and 62 in the low-CCI group.The cohort had a broad range of physical and mental comorbidities, with malignancy making the greatest contribution to mortality. Among these:

• Non-pancreatic neoplasms such as leukemia (n=6, 1.4%), lymphoma (n=21, 4.8%), and solid neoplasms in either localized (n=90, 20.5%) or metastatic stages (n=24, 5.5%)
• Diabetes mellitus, including 42 (9.5%) with and 95 (21.6%) without related complications
• Cardiovascular comorbidities: myocardial infarction (n=23, 5.2%), peripheral vascular disease (n=16, 3.2%), and congestive heart failure (n=34, 7.7%)
• Cerebrovascular disease (n=21, 4.8%)
• Dementia (n=16, 3.6%)
• Liver disease (n=45, 10.2%)

Over median follow-up of 56 months, the investigators observed 12 progressions in worrisome features, two disease-specific deaths, 42 extra-pancreatic deaths, and 44 deaths overall.

The authors concluded that the role of surveillance in individuals with poor overall survival determined through validated scores should be discussed with patients, especially in light of procedural risks, perceived cancer risks, procedural distress, and health resource expenditure. “Prospective utilization of a validated scoring system will enable accurate risk-to-benefit assessment for surveillance interventions and judicious health resource allocation,” they wrote.

Asked for his perspective on the study, Nirav C. Thosani, MD, MHA, of McGovern Medical School at UTHealth in Houston, commended it for highlighting the importance of considering overall comorbidities before subjecting elderly patients to pancreatic cyst surveillance.

“The results clearly support the conclusion that comorbidity assessment tools such as CCI should be included in shared decision-making for surveillance in pancreatic cysts, especially low-risk cysts without worrisome features or high-risk stigmata,” said Thosani, who was not involved in the research.

Mucinous pancreatic cysts do carry potential for progression to malignancy and surveillance is recommended for patients with suspected mucinous pancreatic cysts, he continued. “Despite significant advances in various diagnostic modalities, pre-operative accurate diagnosis remains challenging for these patients. Many are elderly and have significant comorbidities, but despite that undergo various procedures for surveillance given the significant anxiety and concern for development of pancreatic cancer.”

In line with the Yale findings, overall comorbidities are routinely considered at his center before surveillance of this patient begins, he said.

The study had several limitations, the authors acknowledged, including its low number of disease-specific deaths owing to limited follow-up. There was also the possibility that some cases were missed, and since many cysts lacked observable ductal communication or mucinous cytology, they were classified with a presumptive diagnosis.

Furthermore, the study’s association with a regional cancer and liver disease referral program caused a higher comorbidity burden of liver and cancer patients, which does not reflect everyday practice in many other clinical centers. Comorbidity assessment through age-CCI also has some disadvantages since it does not include psychiatric comorbidities, substance use, and arrhythmias.